Evaluation of Students’ Flow State in an E-learning Environment Through Activity and Performance Using Deep Learning Techniques

Abstract

Estimating the flow state of students in a course allows evaluating their sentimental state and the challenges they are facing. In e-learning platforms, the evaluation of flow state is a complex task because it depends on the ability to extract the parameters that better reflect the activity and effort of students. In this scope, the current study proposes a method based on flow theory aiming to provide information about the students' flow state in a course that is taught in an e-learning environment. First, the interaction of students with an e-learning platform that comprises classical e-learning pages and a timeline tool is analyzed, using activity heatmaps and deep neural networks. Then, by taking also in account their grades, the flow state of students is calculated. The resulted data are validated with a statistical analysis that also utilizes student surveys. In order to guarantee that this method is applicable to various profiles, students from different faculties participated in this study. In a period where education is rapidly adapting to online lectures and e-learning platforms, the estimation of student's flow state in e-learning environments can provide useful feedback and data to students and educators.